Standard tests for the cointegrating rank of a vector autoregressive (VAR) process have nonstandard limiting distributions which depend on the characteristics of intercept terms and time trends in the system. In practice these characteristics are often unknown. Therefore modified tests are proposed with limiting distributions which do not depend on the characteristics of deterministic terms under the null hypothesis. One type of tests makes use of lag augmentation, that is, a VAR process of order p + 1 is fitted when the true order is p while the tests are based on the coefficient matrices of the first p lags only. It is shown that Ø 2 limiting distributions are obtained in this way. The price for this simplicity will be reduced power, however. Therefore, we also consider LM (Lagrange multiplier) type tests. These tests are shown to have nonstandard limiting distributions which do not depend on deterministic terms and have better local power than competing LR (likelihood ratio) test...

Following Bierens (1997a,b) and Vogelsang (1998a,b), unit root tests can be constructed which are asymptotically invariant to parameters involved by the short run dynamics of the process. Such an approach is called nonparametric by Bierens (1997b) and can be used to test a wide range of nonlinear models and is robust against structural changes in the short-run dynamics. A variance ratio statistic is suggested which is similar to the test statistic suggested by Kwiatkowski et al. (1992) but assumes nonstationarity under the null hypothesis. A straightforward generalization of the variance ratio statistic is suggested, which can be used to test the cointegration rank in the spirit of Johansen (1988).

In this paper we explore the dynamics of US dollar excess foreign exchange returns for the G10 currencies and the Swiss franc, 1976-97. The non-linear framework adopted is justi¢ed by the results of linearity tests and a parametric bootstrap likelihood ratio statistic which indicate threshold effects or differential adjustment to small and large excess returns. Impulse response analysis suggests that the effect of small shocks to excess returns inside the no-arbitrage band exhibits most persistence. Large shocks outside the band decay most rapidly and also exhibit overshooting. These phenomena are explained in terms of noise trading strategies and transaction costs.

The literature on systems cointegration tests is reviewed and the various sets of assumptions for the asymptotic validity of the tests are compared within a general unifying framework. The comparison includes likelihood ratio tests, Lagrange multiplier and Wald type tests, lag augmentation tests, tests based on canonical correlations, the Stock-Watson tests and Bierens' nonparametric tests. Asymptotic results regarding the power of these tests and previous small sample simulation studies are discussed. Further issues and proposals in the context of systems cointegration tests are also considered briefly. New simulations are presented to compare the tests under uniform conditions. Special emphasis is given to the sensitivity of the test performance with respect to the trending properties of the DGP. Keywords: systems cointegration tests, LR tests, nonparametric tests, asymptotic power, small sample simulations 1 We are grateful to Christian Muller for helping with the computations and ...

This paper provides an extensive Monte-Carlo comparison of several contemporary cointegration tests. Apart from the familiar Gaussian based tests of Johansen, we also consider tests based on nonGaussian quasi-likelihoods. Moreover, we compare the performance of these parametric tests with tests that estimate the score function from the data using either kernel estimation or semi-nonparametric density approximations. The comparison is completed with a fully nonparametric cointegration test. In small samples, the overall performance of the semi-nonparametric approach appears best in terms of size and power. The main cost of the semi-nonparametric approach is the increased computation time. In large samples and for heavily skewed or multimodal distributions, the kernel based adaptive method dominates. For near-Gaussian distributions, however, the semi-nonparametric approach is preferable again. Key words: cointegration testing, adaptive estimation, nonparametrics, semi-nonp...

The short-run dynamics of German mark and US dollar real exchange rates are investigated for a panel of 19 OECD economies in a vector error correction framework for the 1973-96 period. The novel persistence pro¢les approach of Pesaran and Shin (`Cointegration and Speed of Convergence to Equilibrium', Journal of Econometrics, Vol. 71, (1996), pp. 117-143) indicates that the e¡ect of system-wide shocks declines rapidly initially but decays slowly thereafter. It yields an average of just one year for the half-life of such shocks but some seven years before they fully dissipate. These half-life estimates are just one-quarter of the consensus estimates. Our results are consistent with non-linear adjustment and with monetary factors being the main source of real exchange rate volatility. Interest has recently shifted from testing long-run purchasing power parity (PPP) or real exchange rate stationarity to trying to measure the speed of

In this paper we investigate the merits of artificial neural networks in forecasting foreign exchange rates. From previous research it is known that it is hard to beat the random walk model using structural exchange rate models. In this paper we show that by using a suitable multivariate specification a structural model can be derived that beats the random walk. By introducing a new method for multivariate nonlinear cointegration analysis, based on the linear method of Johansen (1988), we construct a neural network error correction model for the yen/dollar, pound/dollar and DM/dollar exchange rates that significantly outperforms both the random walk model and a linear vector error correction model. 1 Introduction Forecasting exchange rates still is a hot topic in modern financial econometrics. Numerous authors have tried a broad range of advanced econometric techniques but mostly without spectacular success. Since the publication of the paper by Meese and Rogoff (1983), it ha...

Abstract not found

This paper considers the specification error tests for omitted nonlinearity in the long-run relationship of the vector error correction model with a null of the standard error correction model. We develop the test statistics for neglected nonlinearity by adding nonlinear functions of the regressors to the long-run relationship. Our test statistics can be calculated using the standard error correction model and its parameter estimates. The test statistics are shown to have the asymptotic standard distribution, and thus the proposed specification error tests can be carried out without inferential difficulty. The alternative distribution of the specification error tests is explored under the local drift and the smooth functional form. The Monte Carlo simulation reveals that the proposed tests have moderate finite sample performance in detecting nonlinear functional form. An economic application of the stock price-dividend relation is provided.

This study builds on the MBA project work of one of the authors, Hanin Smahta, and the supervisory work of the first two authors. The authors acknowledge the financial support of the American University of Beirut Research Board (URB) for funding data collection. All errors are the authors’. 1 Globalization and Investment Opportunities: a Cointegration Study of Arab, U.S. and Emerging Stock Markets There is a debate in the investment circles on whether international diversification is still possible despite growing globalization and the consequent integration between various stock markets all over the world. This study explores whether Arab markets can offer international investors unique risk and return characteristics to diversify international and regional portfolios. A test of co-integration (measuring long-run relationship) is conducted among Arab stock markets, US and emerging markets general indices for a 65-month period (1997-